138 ART. 3. — -B. KOTü : 



The writer was probably tbe first amongst specialists, if he 

 may be called so, who hurried to the scene and viewed the 

 front or west side of the grand volcanic display, without be- 

 ing able to come to any definite idea as to the i-elation of vents 

 and geologic structure ; but on seeing the lineal arrangement 

 (Text-fig. 19 ; PL XL Figs. 1-2) of bocche on the rear, the Nabé- 

 yama side, the writer immediately conceived the existence of a 

 mountain fracture extending on both flanks of the volcano. 



During preliminary work to learn the distribution of the various 

 lava flows, both ancient and historic, and also the history of the 

 building-up of the volcano of Sakura-jima, the writer's view, 

 however, became somewhat modified. Though apparently a simple 

 overtowering konide, Sakura-jima is really a triple one, piled on 

 the shoulder one after the other in a meridional direction in the 

 order from the north top, then the south, and finally the middle 

 (Text-fig. 8, and PL III. Fig. 2). 



The constructive mature age of vulcanicity having passed 

 away, the declining period of pericentric flank eruptions was 

 ushered in in historic times, when lava poured forth around and 

 outside of top-craters all round the slope after the fashion of the 

 petals of a flower, firmly sheeting the old composite konide with a 

 lava coat of mail. 



There are, however, portions of the slope still remaining free of 

 lava flows (Geologic Map), and these intervening gaps — tJ/e j^ositions 

 of comparatire weakness in the superstriiciure of the volcano — icere 

 selected for the effusion of hot magma in the recent eruption. More- 

 over, the positions of vents lie at the overlapping junction-line of 

 the north and the south cones which are the dominant topographic 

 elements of the composite triple konide of Sakm-a-jima. 



The préexistence of a transverse fracture is universally ad- 



